the state of serverless computing
play

The State of Serverless Computing or, Fixing - PowerPoint PPT Presentation

Jul 13, 2023 •49 likes •549 views

The State of Serverless Computing or, Fixing Dysfunction-as-a-Service Chenggang Wu RISE Lab, UC Berkeley QCon New York 06/24/2019 UC Berkeley: CS 61A Fall 2018 The State of Serverless Computing QCon New York 06/24/2019 The State of

  1. The State of Serverless Computing or, Fixing Dysfunction-as-a-Service Chenggang Wu RISE Lab, UC Berkeley QCon New York 06/24/2019

  2. UC Berkeley: CS 61A Fall 2018 The State of Serverless Computing QCon New York 06/24/2019

  3. The State of Serverless Computing The State of Serverless Computing QCon New York 06/24/2019 QCon New York 06/24/2019

  4. Making Programmers Productive Key Question: Where will code be run? In the ☁ ! The State of Serverless Computing QCon New York 06/24/2019

  5. Background: Serverless Computing The State of Serverless Computing QCon New York 06/24/2019

  6. What is serverless computing? Serverless computing is a programming abstraction that enables users to upload programs, run them at any scale, and pay only for resources used. The State of Serverless Computing QCon New York 06/24/2019

  7. Functions-as-a-Service (FaaS) • AWS Lambda, Google Cloud Functions, OpenWhisk (IBM), Azure Functions, OpenLambda, OpenFaaS, kNative… • Optimized for simplicity – register functions, enable triggers, and scale transparently The State of Serverless Computing QCon New York 06/24/2019

  8. Academic Interest in Serverless The State of Serverless Computing Craft Conference 2019 – Budapest, Hungary

  9. Industrial Interest in Serverless The State of Serverless Computing Craft Conference 2019 – Budapest, Hungary

  10. What is FaaS good at today? f() f() f() f() Embarrassingly parallel tasks Workflow orchestration The State of Serverless Computing QCon New York 06/24/2019

  11. Wait! What about… • There are other serverless services, too! • e.g., Google Cloud Dataflow, AWS Athena, Snowflake… • Often referred to as Backend-as-a-Service (BaaS) We’re primarily interested in generality. The State of Serverless Computing QCon New York 06/24/2019

  12. Limitations on FaaS Today Limited execution lifetimes x No inbound network connections IO is a bottleneck No specialized hardware The State of Serverless Computing QCon New York 06/24/2019

  13. But that’s okay: Everything is functional! • Functional programs don’t have side effects or mutable state! • And it is called AWS Lambda The State of Serverless Computing QCon New York 06/24/2019

  14. Dysfunction-as-a-Service • FaaS is not designed for functional programming because real applications share state • FaaS is poorly suited for all of these The State of Serverless Computing QCon New York 06/24/2019

  15. Quantifying The Pain of FaaS How FaaS Disappoints Famous Computer Scientists The State of Serverless Computing QCon New York 06/24/2019

  16. Even Functional Programming is Slow! 1000 Latency (ms) 737 573 285 239 100 10 1 Lambda Lambda Lambda AWS (S3) (Dynamo) (Direct) Step FNs Median and 99 th percentile latencies for composing two arithmetic functions on AWS Lambda. The State of Serverless Computing QCon New York 06/24/2019

  17. Even Functional Programming is Slow! 1000 Latency (ms) 737 573 178 285 239 100 79.6 10 1 Lambda Lambda Lambda AWS (S3) (Dynamo) (Direct) Step FNs Median and 99 th percentile latencies for composing two arithmetic functions on AWS Lambda. The State of Serverless Computing QCon New York 06/24/2019

  18. Even Functional Programming is Slow! 3346 1000 Latency (ms) 737 573 569 178 285 239 100 79.6 10 1 Lambda Lambda Lambda AWS (S3) (Dynamo) (Direct) Step Fns Median and 99 th percentile latencies for composing two arithmetic functions on AWS Lambda. The State of Serverless Computing QCon New York 06/24/2019

  19. Shared Mutable State The State of Serverless Computing QCon New York 06/24/2019

  20. Shared Mutable Storage The State of Serverless Computing QCon New York 06/24/2019

  21. Ideal Serverless Storage Autoscaling Low Latency The State of Serverless Computing QCon New York 06/24/2019

  22. (In)Consistency Guarantees ? Shared Counter The State of Serverless Computing QCon New York 06/24/2019

  23. No Inbound Network Connections Enables Process Easy Fault Migration Tolerance The State of Serverless Computing QCon New York 06/24/2019

  24. Indirect Communication Write Read The State of Serverless Computing QCon New York 06/24/2019

  25. We can fix that! The State of Serverless Computing QCon New York 06/24/2019

  26. A Platform for Stateful Serverless Computing The State of Serverless Computing QCon New York 06/24/2019

  27. Background: Anna • High performance across orders of magnitude in scale ü 10x faster than Redis/Cassandra in a geo-distributed deployment • Autoscaling & cost-efficient ü 500x faster than Amazon DynamoDB for the same cost The State of Serverless Computing Chenggang Wu, Jose Faleiro, Yihan Lin, and Joseph M. Hellerstein. "Anna: A KVS for Any Scale." IEEE Transactions on Knowledge and Data Engineering (2019). QCon New York 06/24/2019 Chenggang Wu, Vikram Sreekanti, and Joseph M. Hellerstein. "Autoscaling Tiered Cloud Storage in Anna." Proceedings of the VLDB Endowment 12, no. 6 (2019): 624-638.

  28. Fluent: FaaS-over-Anna • Maintain disaggregation of compute & state • Make serverless a viable option for stateful applications • Use Anna for both storage and communication Anna The State of Serverless Computing QCon New York 06/24/2019

  29. Fluent: FaaS-over-Anna Network Boundary Anna The State of Serverless Computing QCon New York 06/24/2019

  30. Logical disaggregation with physical colocation The State of Serverless Computing QCon New York 06/24/2019

  31. Fluent: FaaS-over-Anna Network Boundary Anna The State of Serverless Computing QCon New York 06/24/2019

  32. Key Idea: Caching • Enable low-latency data access by caching data close to code execution • Communication (and composition) is achieved via a fast-path on top of KVS puts and gets The State of Serverless Computing QCon New York 06/24/2019

  33. Challenge: Cache Consistency • tl;dr: we can provide a variety of coordination-avoiding consistency modes – which is better than S3 or DynamoDB! • This is done by encapsulating program state in lattices The State of Serverless Computing QCon New York 06/24/2019

  34. Lattice • Data structure that accepts incoming update in a way that is associative, commutative, and idempotent (ACI). • Achieves eventual replica convergence The State of Serverless Computing QCon New York 06/24/2019

  35. Causal Consistency • Strongest consistency level that doesn’t require coordination • Causally-related updates will be revealed in an order that respects causality • In addition, guarantee • Repeatable read • Atomic visibility The State of Serverless Computing QCon New York 06/24/2019

  36. Function Composition, Revisited 3346 1000 Latency (ms) 737 573 569 178 285 239 100 79.6 10 6.37 3.59 1 Fluent Lambda Lambda Lambda AWS (S3) (Dynamo) (Direct) Step Fns Median and 99 th percentile latencies for composing two arithmetic functions on AWS Lambda and Fluent. The State of Serverless Computing QCon New York 06/24/2019

  37. Case Study: Prediction Serving The State of Serverless Computing QCon New York 06/24/2019

  38. Prediction Serving • Generate predictions from pretrained machine learning models Model Replica Model Join with Combine Clean Input Reference Data Replica Results Model Replica • At first blush, a great fit for serverless infrastructure The State of Serverless Computing QCon New York 06/24/2019

  39. The State of Serverless Computing QCon New York 06/24/2019

  40. Background: SqueezeNet • State-of-the-art image classification model (developed at Berkeley!) The State of Serverless Computing QCon New York 06/24/2019

  41. Prediction Serving 400 361.7 350 Latency (ms) 300 282.2 250 200 156.3 150 153.8 122.9 100 90.6 50 0 Fluent Python AWS SageMaker Median and 99 th percentile latencies for SqueezeNet on Fluent and AWS SageMaker. The State of Serverless Computing QCon New York 06/24/2019

  42. The Future of Cloud Programming The State of Serverless Computing QCon New York 06/24/2019

  43. Looking Back: Disappointed Computer Scientists • Functional programming is slow • Communication through slow storage • Poor consistency guarantees The State of Serverless Computing QCon New York 06/24/2019

  44. Making FaaS Functional • Embrace state • Easy things become better • Hard things become easy A step on our road towards a programmable cloud. The State of Serverless Computing QCon New York 06/24/2019

  45. Our Vision • Serverless will change the way that we write software and the way that programming infrastructure works • Cloud-native programming models • Enable users to take advantage of millions of cores and petabytes of RAM The State of Serverless Computing QCon New York 06/24/2019

  46. Moving Forward from FaaS Building Developer Tools The State of Serverless Computing QCon New York 06/24/2019

  47. UC Berkeley: CS 61A Fall 2018 The State of Serverless Computing QCon New York 06/24/2019

  48. Moving Forward from FaaS Building Developing Developer Autoscaling Tools Policy The State of Serverless Computing QCon New York 06/24/2019

  49. Moving Forward from FaaS Designing Building Developing Developer Autoscaling SLOs & SLAs Tools Policy The State of Serverless Computing QCon New York 06/24/2019

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Functions as a Service (Serverless computing) Motivation All require at least one server to

Functions as a Service (Serverless computing) Motivation All require at least one server to

Functions as a Service (Serverless computing) Motivation All require at least one server to be running at all times Want something that costs $0 if not used Portland State University CS 410/510 Internet, Web, and Cloud Systems Serverless

1.3k views • 52 slides
Stateful Serverless Sean Walsh @SeanWalshEsq We predict that Serverless Computing will grow

Stateful Serverless Sean Walsh @SeanWalshEsq We predict that Serverless Computing will grow

Towards Stateful Serverless Sean Walsh @SeanWalshEsq We predict that Serverless Computing will grow to dominate the future of Cloud Computing. - Berkeley CS Department Cloud computing simplified: a Berkeley view on serverless computing

748 views • 40 slides
FaaS You Like It! @ewanslater Serverless CNCF Definition Serverless computing refers to

FaaS You Like It! @ewanslater Serverless CNCF Definition Serverless computing refers to

FaaS You Like It! @ewanslater Serverless CNCF Definition Serverless computing refers to the concept of building and running applications that do not require server management . It describes a finer-grained deployment model where

613 views • 45 slides
Kotlin Serverless Framework Vladislav Tankov What is serverless? cloud-computing execution model

Kotlin Serverless Framework Vladislav Tankov What is serverless? cloud-computing execution model

Kotlin Serverless Framework Vladislav Tankov What is serverless? cloud-computing execution model , in which the cloud provider runs the server and dynamically manages the allocation of machine resources Wikipedia How are serverless

995 views • 46 slides
Serverless On Your Own Terms Using Knative Context Serverless more than Function Serverless

Serverless On Your Own Terms Using Knative Context Serverless more than Function Serverless

Serverless at Google @mchmarny Serverless On Your Own Terms Using Knative Context Serverless more than Function Serverless Models Operator No Infra Management Managed Security Pay only for usage Developer Service-based Event-driven Open

1.06k views • 51 slides
F AASM : Lightweight Isolation for Efficient Stateful Serverless Computing Simon Shillaker and

F AASM : Lightweight Isolation for Efficient Stateful Serverless Computing Simon Shillaker and

F AASM : Lightweight Isolation for Efficient Stateful Serverless Computing Simon Shillaker and Peter Pietzuch Large-scale Data and Systems Group, Imperial College London Serverless Big Data Vision Serverless functions Application Big data

431 views • 30 slides
Serverless Performance on a Budget Erwin van Eyk The central trade-off in serverless computing

Serverless Performance on a Budget Erwin van Eyk The central trade-off in serverless computing

Serverless Performance on a Budget Erwin van Eyk The central trade-off in serverless computing High Performance Infinite scaling High availability Low latency 2 The central trade-off in serverless computing High Performance Low

936 views • 77 slides
cloudstate.io serverless 2.0 with cloudstate Sean Walsh | Field CTO and Cloud Evangelist @

cloudstate.io serverless 2.0 with cloudstate Sean Walsh | Field CTO and Cloud Evangelist @

cloudstate.io serverless 2.0 with cloudstate Sean Walsh | Field CTO and Cloud Evangelist @ Lightbend We predict that serverless computing will grow to dominate the future of cloud computing. Berkely CS Department why serverless 2.0?

701 views • 38 slides
In Search of a Fast and Efficient Serverless DAG Engine Benjamin Carver, Jingyuan Zhang, Ao

In Search of a Fast and Efficient Serverless DAG Engine Benjamin Carver, Jingyuan Zhang, Ao

In Search of a Fast and Efficient Serverless DAG Engine Benjamin Carver, Jingyuan Zhang, Ao Wang, Yue Cheng Serverless Computing Emerging cloud computing platform based on the composition of fine-grained user-defined functions

832 views • 31 slides
How Serverless Changes the IT Department Paul Johnston Opinionated Serverless Person

How Serverless Changes the IT Department Paul Johnston Opinionated Serverless Person

How Serverless Changes the IT Department Paul Johnston Opinionated Serverless Person Serverless all the things Medium/Twitter: @PaulDJohnston Velocify Conf London 2018 Paul Johnston Experienced Interim CTO and Serverless Consultant

1.11k views • 63 slides
Serverless networking (peer-to-peer computing) Peer-to-peer models Client-server computing

Serverless networking (peer-to-peer computing) Peer-to-peer models Client-server computing

5/7/08 Serverless networking (peer-to-peer computing) Peer-to-peer models Client-server computing servers provide special services to clients clients request service from a server Pure peer-peer computing all systems have equivalent

586 views • 20 slides
An Investigation of the Impact of Language Runtime on the Performance and Cost of Serverless

An Investigation of the Impact of Language Runtime on the Performance and Cost of Serverless

An Investigation of the Impact of Language Runtime on the Performance and Cost of Serverless Functions David Jackson & Gary Clynch Institute of Technology, Tallaght 4th International Workshop on Serverless Computing (WoSC4) 20th December

568 views • 35 slides
Serverless Gardens IoT + Serverless johncmckim.me twitter.com/@johncmckim

Serverless Gardens IoT + Serverless johncmckim.me twitter.com/@johncmckim

Serverless Gardens IoT + Serverless johncmckim.me twitter.com/@johncmckim medium.com/@johncmckim John McKim Software Engineer at A Cloud Guru Contribute to Serverless Framework @johncmckim https://acloud.guru Serverless Framework

814 views • 50 slides
Kotless Kotlin Serverless Framework Vladislav Tankov @vdtankov October 15, 2020 Introduction

Kotless Kotlin Serverless Framework Vladislav Tankov @vdtankov October 15, 2020 Introduction

Kotlin 1.4 Online Event Kotless Kotlin Serverless Framework Vladislav Tankov @vdtankov October 15, 2020 Introduction to serverless A few words on the hottest topic Serverless is a cloud-computing execution model in which the cloud

884 views • 33 slides
SERVERLESS COMPUTING FOR DATA-PROCESSING ACROSS PUBLIC AND FEDERATED CLOUDS Sebastin Risco,

SERVERLESS COMPUTING FOR DATA-PROCESSING ACROSS PUBLIC AND FEDERATED CLOUDS Sebastin Risco,

Instituto de Instrumentacin para Imagen Molecular Universitat Politcnica de Valncia Spain SERVERLESS COMPUTING FOR DATA-PROCESSING ACROSS PUBLIC AND FEDERATED CLOUDS Sebastin Risco, Alfonso Prez, Miguel Caballer, Germn Molt

218 views • 21 slides
Centralized Core-granular Scheduling for Serverless Functions Kostis Kaffes , Neeraja J.

Centralized Core-granular Scheduling for Serverless Functions Kostis Kaffes , Neeraja J.

Centralized Core-granular Scheduling for Serverless Functions Kostis Kaffes , Neeraja J. Yadwadkar, Christos Kozyrakis Serverless Computing is Convenient for Users Users: Define a function Specify events as execution triggers Pay only

364 views • 20 slides
Multi-Task Learning and Matrix Regularization Andreas Argyriou Department of Computer Science

Multi-Task Learning and Matrix Regularization Andreas Argyriou Department of Computer Science

Multi-Task Learning and Matrix Regularization Andreas Argyriou Department of Computer Science University College London Collaborators T. Evgeniou (INSEAD) R. Hauser (University of Oxford) M. Herbster (University College London)

910 views • 54 slides
1 Shaadi by Marriott 2 There is no greater or more prestigious venue to host events with

1 Shaadi by Marriott 2 There is no greater or more prestigious venue to host events with

1 Shaadi by Marriott 2 There is no greater or more prestigious venue to host events with stature. Whether large or small, opulent or intimate, our dedicated and passionate team can create a memorable experience 3 Congratulations, Thank

448 views • 30 slides
PT Kawasan Industri Jababeka Tbk. Investor Presentation September 2019 0 Leading township

PT Kawasan Industri Jababeka Tbk. Investor Presentation September 2019 0 Leading township

PT Kawasan Industri Jababeka Tbk. Investor Presentation September 2019 0 Leading township developer & infrastructure powerhouse PT Kawasan Industri Jababeka Tbk. ("KIJA") is a leading township developer with an established track

686 views • 30 slides
Internet content HTML SGML CSS XML XHTML MIME HTTP DD1335 (Lecture 2) Basic Internet

Internet content HTML SGML CSS XML XHTML MIME HTTP DD1335 (Lecture 2) Basic Internet

Internet Content Internet content HTML SGML CSS XML XHTML MIME HTTP DD1335 (Lecture 2) Basic Internet Programming Spring 2010 1 / 27 Internet Content Objectives What HTML is, what its origins are, and where one can find information

1.04k views • 101 slides
MATE TERNAL MORTALITY REDUCT CTION IN IN COVID ID PDG Dr . HIMANSU BASU, MD, PhD, FRCS,FRCOG

MATE TERNAL MORTALITY REDUCT CTION IN IN COVID ID PDG Dr . HIMANSU BASU, MD, PhD, FRCS,FRCOG

COVID 19 PANDEMIC PROTECTING MOTHERS AND BABIES DURING PREVENTABLE THE PANDEMIC AND BEYOND MATE TERNAL MORTALITY REDUCT CTION IN IN COVID ID PDG Dr . HIMANSU BASU, MD, PhD, FRCS,FRCOG PANDEMIC ERA & Consultant

309 views • 28 slides
Business Model Canvas Business Model : Definition A business model describes how an idea will

Business Model Canvas Business Model : Definition A business model describes how an idea will

Business Model Canvas Business Model : Definition A business model describes how an idea will create value. How do they make money? Product & service Subscription to a Ad revenue sales service How do they make money? Commission

1.04k views • 23 slides
Guidelines for Biodiversity Management Committees (BMCs) National Biodiversity Authority

Guidelines for Biodiversity Management Committees (BMCs) National Biodiversity Authority

Guidelines for Biodiversity Management Committees (BMCs) National Biodiversity Authority Formation of BMC Section 41 of the Biological Diversity Act states as follows:- Every local body shall constitute a Biodiversity Management Committee

532 views • 23 slides
Running Successful Climbing Competitions Robert Adie BMC Competitions Officer Paul Twomey The

Running Successful Climbing Competitions Robert Adie BMC Competitions Officer Paul Twomey The

Running Successful Climbing Competitions Robert Adie BMC Competitions Officer Paul Twomey The Climbing Academy Tom Vaessen Reading Climbing Centre Robert Adie BMC Competitions Officer National and International competitions

329 views • 12 slides

More recommend